[prev in list] [next in list] [prev in thread] [next in thread] 

List:       python-csv
Subject:    [Csv] CSV ignores lineterminator
From:       andrewm () object-craft ! com ! au (Andrew McNamara)
Date:       2004-04-06 1:49:27
Message-ID: 20040406014927.7CB253C744 () coffee ! object-craft ! com ! au
[Download RAW message or body]

>I understand the difficulty of these sorts of tradeoffs.  My main wish is 
>that the documentation was clearer so that I could have avoided spending so 
>much time trying to understand what was going wrong.

I'd agree with that. Part of the problem is time - this gets done in time
we steal from other projects, and that's in short supply at the moment. 

Of course, if you're annoyed enough at the state of the doco, maybe you'd
consider giving us a hand with them? Both the PEP and the python module
doco could do with a thorough going over before Python 2.4 comes out. 8-)

>Speaking of which -- and I don't intend to nag you indefinitely -- the
>documentation says that the dialect argument to reader or writer takes a
>string that identifies a dialect previously registered using
>register_dialect.  However, the email I got from Skip Montanaro passed a
>dialect directly -- and that seems to work too.  Is it permitted to use
>either a string or a dialect?

If I remember correctly, the underlying C module will accept either
a dialect object, or keyword arguments describing a dialect (which,
I think, it converts to a dialect object). I think you can actually
mix dialect and keyword arguments to get an effect like "I want excel,
but with these changes". That was the intention, anyway... 8-)

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/

[prev in list] [next in list] [prev in thread] [next in thread]